The Correlates of Voter Turnout in the United Kingdom

¶ … Unemployment on Voter Turnout Rates in Britain's Elections

In many countries where voting is a right but not compulsory, voter turnout levels have declined precipitously in recent years, causing a growing number of political scientists to examine the effects of various variables on these rates. By contrast, in countries where voting is compulsory such as Australia, Belgium and Italy, voter turnout levels approach 100% (Forman, 2002). The research on political behavior to date has analyzed a wide range of variables that affect voter turnout positively (such as mail ballots or electronic voting from home) or negatively (such as poll taxes or voter identification requirements) (Childers & Binder, 2012). There has been far less research, though, concerning the potential impact of other variables such as employment status on voter turnout levels. For example, Incantalupo (2015) emphasizes that, "At present, gaps exist in our understanding of how unemployment affects political attitudes and behavior in part because there has been so little attention paid to unemployment as a personal hardship and the disparate contexts in which someone can experience unemployment" (p. 4). In fact, it has been more than 30 years when unemployment levels were high since scholarship was focused on unemployment and theories of unemployment were created and evaluated (Incantalupo, 2015), a gap that the proposed study will help fill as discussed further below.

Background and Overview

The importance of voter turnout has become the focus of a growing body of research as a result of differing rates of political participation by various economic and political groups (Schur & Shields, 2002). This research has built on an original study by Downs in 1957 that described the expected utility theory of voting as affecting voter turnout levels. According to Nicholson and Miller (1993), "In its original form, this theory posits that decisions to vote hinge on two factors: the probability that a single vote will affect the outcome of an election and the costs associated with voting" (p. 199). Since the likelihood that a single vote will have an impact on the outcome of an election, many citizens may not vote because the benefits of their voting are discounted compared to the costs that are involved (Nicholson & Miller, 1993).

From Downs' (1957) perspective, it is frequently more "rational" for citizens to not vote because they will have little impact on an election's outcome. In this regard, Abramson and Diskin (2007) note that, "Although the cost of voting is low, the probability that a single voter will affect the outcome in a large election is negligible. Rational citizens, realizing that their probability of affecting the outcome is negligible, may choose to abstain" (p. 501). Downs (1957) conceptualized his theory with the following equation:

R = (B)(P) -- C + D

In this equation, R is the total reward a citizen will gain after voting, B is the benefit one thinks he or she will accrue from having his or her preferred candidate win in an election, P is an individual's perception of the probability that his or her vote will change the outcome of an election, C is the cost associated with voting in terms of time, money and other resources, while D is the psychic satisfaction an individual would gain from voting. If R is positive, the individual is likely to gain a reward from voting that outweighs costs. Hence, the more positive R is, the more likely citizens are to vote.

According to Harder and Krosnick (2008), this equation set forth above illustrates the so-called "paradox of voting" (p. 526). The paradox of voting holds that, "Voting yields benefits only when supported by collective action, so most people should never pay the costs because their effort will never ensure the acquisition of benefits" (Harder & Krosnick, 2008, p. 526).

The results of a study by Incantalupo (2015) identified a relationship between unemployment and voter turnout. Based on his analysis of the effects of unemployment on voter turnout, Incantalupo (2015) reports that, "The economic and social contexts in which [voters] experience unemployment structure the ways in which their economic situations influence their political behavior" (p. 3). Interestingly, the results of this study identified a negative relationship on voter turnout levels and low unemployment rates and a positive relationship between voter turnout levels and high unemployment levels (Incantalupo, 2015). In this regard, Incantalupo (2015) adds that, "Broadly speaking, personal experience with involuntary unemployment is a mobilizing phenomenon that increases voter turnout in high-unemployment contexts and a demobilizing phenomenon that decreases voter turnout in low-unemployment contexts" (p. 3).

Voter turnout levels are even lower when people are unemployed and disabled. For instance, according to Schur and Shields (2002), "The lower turnout is concentrated among people with disabilities who are not employed" (p. 167). These findings indicate that unemployment can adversely affect voter turnout levels due to decreased social capital and identification with mainstream society (Schur & Todd, 2002). In addition, Schur and Shields (2002) note that, "The findings also support the idea that general mobility and major life transitions can be important influences on voter turnout in general, and raise questions on the causal relations among age, employment, efficacy, and voter turnout that should be a focus of future research" (p. 168). Some of the other salient demographic factors that have been shown to affect voter turnout levels include the following:

Age -- People are more likely to vote as they progress from early adulthood (early 20s) through middle adulthood (mid 30s and 40s). However, after about the age of 75, voter participation decreases. This is due to the possibility that older citizens are more established in their communities, able to have more free time and are generally more interested in political outcomes compared to younger citizens. On the other hand, people aged 75 and above may vote less due to declines in physical health, mobility and energy levels.

Gender -- Though fewer women took part in the electoral process years ago than they currently do, women today have voted at the same rate as men and sometimes at even higher rates, due to increased political interests.

Education level -- People with more formal education are more likely to vote, due to the fact that education may impart the skills which enhance a person's ability to understand how the civic process works and how to navigate the requirements of registration. Education could also encourage people to vote by instilling civic duty, forcing them to become interested in the political process.

Race - Whites tend to vote at higher rates compared to other racial groups. Voter turnout rates among African-Americans are lower than that of whites, but has only increased recently due to relaxed discriminatory voter registration laws. They could have been inspired to vote as a result of dissatisfaction with their minority status. Asian-Americans also have lower voter turnout rates than whites, but Chinese, Vietnamese and Korean-Americans in particular have low registration and voter turnout rates, due to differences in the likelihood of registering.

Income - Wealthier people vote at higher rates. Less wealthy people have less time and resources available to learn about the electoral process when compared to wealthy people. Could also be due to the fact that more wealthy people perceive that they have a greater interest at stake in elections compared to the less wealthy.

Marital Status -- Married couples tend to vote at higher rates than single-parent households possibly because politically motivated people are able to inspire less motivated spouses, either through persuasion, or simply by exposing the spouse to political information. Such would not be the case in a single-parent household. Divorce greatly increases voter turnout rates since divorced people now have more free time to devote to learning about elections and politics in general.

Civic Duty -- Some people believe they have a moral obligation to participate in politics and are therefore more inclined to take part in elections. Similarly, those who believe that all citizens have the obligation to vote go to the polls more often than those who do not hold such a belief.

Trust -- People who can trust others are more likely to vote in an election. Distrustful people would think of the political system as corrupt, discouraging them from taking part in elections. However, there are exceptions when low levels of trust might inspire some to vote only if it minimizes the damage they might fear others could cause.

Voting in the United Kingdom

In the UK's parliamentary election, registered voters in every area of the country vote for a Member of Parliament (MP) to represent them in the House of Commons. The House of Commons is an elected body consisting of 650 members known as Members of Parliament (MPs). There are 650 geographical areas, which are called constituencies. An individual can vote in a UK parliamentary general election if that person is registered to vote and is:

Aged 18 years or over on polling day;

A UK citizen, Commonwealth citizen or a citizen of the Irish Republic; and,

Not legally excluded from voting.

The United Kingdom is currently divided into 650 parliamentary constituencies, each of which is represented by one Member of Parliament (MP) in the House of Commons. There are currently:

533 constituencies in England;

59 constituencies in Scotland;

40 constituencies in Wales and 18 in Northern Ireland.

At present, though, fully 82% of the seats in the House of Commons correspond to English constituencies (Boisvert & Smith, 2015). Therefore, voter turnout for House of Commons elections vary from region to region as well as constituency to constituency. For example, according to Abramson and Diskin (2007), "In the United Kingdom's general election of 2005, the mean turnout for all constituencies was 60.9% of voting-age population, but the standard deviation among the constituencies was 6.4, ranging from a low of 37.3% in Staffordshire South to a high of 76.3% in Dorset West" (p. 501).

The dismally low voter turnout level that was experienced in some constituencies during the 2001 general election has become the focus of nationwide efforts. In this regard, Ferguson (2008) notes that, "The United Kingdom began to register an interest in the issue of citizen engagement -- or more precisely, citizen disengagement -- following the 2001 general election" (p. 28). During the 2001 general election, voter turnout was the lowest level (59%) that has been experienced since 1918 (Ferguson, 2008). Likewise, voter turnout levels during 1997 were lower than in 1992, and although there was a modest increase (2%) in general voter turnout levels experienced in 2005, there was a corresponding decrease of 2% among the youngest voters (Ferguson, 2008). Therefore, identifying the correlates of low voter turnout levels represents a timely and valuable enterprise as discussed below.

Purpose of the Study

The purpose of the proposed study will be to examine the effects of unemployment on voter turnout rates in Britain's 2001 and 2010 general elections using 2001 and 2010 unemployment and voter turnout data. The theory that unemployment levels affect voter turnout rates will be tested using data from the Office for National Statistics (ONS) in Britain, while controlling for variables such as age, gender and marital status obtained from the 2001 and 2011 UK census, as well as electoral closeness measured by 2001 and 2010 election results. In this context, the closer the election polls for certain candidates appear to be prior to election day, the more convinced citizens will be that their votes could ultimately determine the election outcome. This would encourage them to cast their votes. However, if pre-election polls suggest that a landslide victory is likely to occur, then voter turnout rates would decline. Likewise, the greater the gap between a person's attitude toward one candidate and his or her attitude toward a competing candidate, the more likely the person will vote. Moreover, if a citizen strongly prefers one candidate over the other, he or she will be more willing to turn out in a given election and vote.

Research Question

The proposed study will be guided by the following research questions:

Does unemployment affect voter turnout rates in Britain's general elections?

If unemployment does affect voter turnout rates, do high unemployment levels result to an increase in voter turnout rates in a particular election?

Does the relationship between unemployment and voter turnout rates differ for each of the 650 UK constituencies while controlling for age, gender, marital status and electoral closeness?

Hypothesis

The proposed study's guiding hypothesis is that unemployment has a definite impact on voter turnout rates in Britain's general elections. More precisely, constituencies experiencing high unemployment rates will eventually see an increase in voter turnout rates. It is expected that the proposed study will find that the relationship between unemployment and voter turnout rates is stronger in the constituencies which show greater proportions of females than males, greater proportions of elderly citizens than younger citizens, greater proportions of married couples than those who are single or divorced, as well as in the constituencies which show a closer percentage share of votes won by the top two political parties than others. In other words, the closer the percentage share of votes received by the top two political parties is, the higher voter turnout is expected to be in each UK constituency.

Theoretical Framework

The majority of the studies to date concerning voter turnout levels have used the rational actor theoretical framework. According to Schur and Shields (2002), "Central to this model is the recognition that the decision to vote, at least for most citizens, is a marginal (low cost / low benefit) action" (p. 168). A study by Nicholson and Miller (1997) used the rational actor model which identified a relationship between the closeness of elections and voter turnout levels. In this regard, Nicholson and Miller report that, "The closeness of elections is a central concern in rational actor models of voter turnout" (p. 200).

Although rational expectations concerning the size or closeness of an election have been identified as affecting voter turnout levels, there are other salient variables involved as well. For instance, Abramson and Diskin (2007) cite seminal research by Merriam and Gosnell (1924), Lipset (1960) and Dahl (1961) who studied political involvement and participation rates. In this regard, Abramson and Diskin report that, "Abundant evidence shows that many of the social variables identified by Lipset and Dahl correlate with turnout. For example, members of lower socioeconomic groups tend to be far less politically involved than those belonging to higher status groups" (2007, p. 501). Likewise, these and other studies have also identified correlations between low voter turnout levels and socioeconomic status (Abramson & Diskin, 2007).

In addition, the rational actor model holds that voter turnout levels depend on the extent to which citizens believe that their vote will make a difference in an election outcome (Nicholson & Miller, 1997). Therefore, it is reasonable to posit that the extent to which individual voters are already disenfranchised from mainstream society due to unemployment may be the extent to which they will be less likely to make the effort to vote. This assertion is congruent with Incantalupo (2015) who reports, "The financial burden of unemployment and the demands and stresses of looking for new work could lower one's likelihood of participation" (p. 5). Citizens who are unemployed may therefore lack the same incentives to vote as their employed counterparts.

Although there remains a dearth of timely and relevant studies concerning the relationship between unemployment and voter turnout levels, the limited amount of research in this area indicates that unemployment can be a demobilizing phenomenon that adversely affects an individual's desire to vote. In this regard, Incantalupo (2015) points out that, "Citizens who experience job loss are unlikely to get sufficiently aroused to make it a point to head to the polls on Election Day. To the extent that citizens perceive job loss as a personal problem with which they must cope and that they must remedy, job loss will not increase their likelihood of voting" (p. 5). While some researchers have suggested that unemployed citizens may have more leisure time to devote to politics, there is a growing consensus that unemployment causes declines in voter turnout levels. From a strictly pragmatic perspective, this just make sense because many unemployed citizens are more concerned about their immediate circumstances and survival than they are about abstract political issues. As Incantalupo concludes, "When a person experiences economic adversity his scarce resources are spend on holding body and soul together -- surviving -- not on remote concerns like politics" (2015, p. 5).

It is important to note, though, that while the rational actor model "might explain why patterns of collective action persist it does not explain how they arise" (Goodin & Klingemann, 1998, p. 157). Consequently, it is also important to take other variables into account in examining voter turnout levels. To this end, the independent variable that will be used in the proposed study will be unemployment, the dependent variable will be voter turnout (e.g., likelihood of voting = motivation to vote * ability to vote/difficulty in voting), and the control variables are age, gender, marital status (categorized into the five groups: single, married, widowed, divorced or in a registered same-sex civil partnership), and electoral closeness (measured by the difference in the percentage share of votes received by the top two political parties in each constituency).

The rationale in support of this approach is that if the motivation and ability an individual has to vote increases, the likelihood that he or she will turn out to vote in a given election increases, however, if there is a great degree of difficulty associated with voting, the likelihood that an individual will turn out to vote in a particular election decreases. Motivation could come from a strong preference for one candidate over a competitor, peer pressure, or generally from the belief that being a responsible citizen requires that a person votes. Ability to vote refers to an individual's capacity to understand and meet requirements in order to be eligible to vote legally. Difficulty of voting refers to the factors outside a voter's control that would restrict him or her from voting. Factors such as the convenience of registration procedures, as well as the distance of the nearest polling station from a person's home would play a role in affecting the likelihood of voter participation.

Methodology

The independent variable will be defined as the constituency unemployment rate (UnempConstRate), while the dependent variable will be defined as the percentage of eligible voters who registered their votes in the general elections. Age will be classified into age groups ranging from 0 to 14 years, 15 to 29 years, 30 to 44 years, 45 to 59 years, 60 to 74 years, 75 to 89 years and 90+, gender will be defined as the number of males and females present in each constituency, marital status will be defined as the number of people who are either single, married, widowed, divorced, or in a registered same-sex civil partnership in each constituency, while electoral closeness will be defined as the difference between the percentage share of votes the top two political parties of each constituency receive in a particular election

In order to determine whether unemployment has an impact on voter turnout, the proposed study will employ a regression with controls on observables such as age, gender, marital status and electoral closeness. There are several factors that tend to determine whether someone votes and all of these factors will be captured in my Ordinary Least Squares (OLS) estimation model. The Ordinary Least Squares Estimation to be used in the proposed study is as follows:

voter turnoutit = 0+ 1 unemplit+ 2 ageit+ 3 genderit+ 4 maritalit+ 5 electoralit+ ai + uit i =, ..., n t = 1, 2

(where ai represents the error term that varies between constituencies, but not over time and uit represents the error term that varies between constituencies and across time). A fixed effect will be used for all 650 UK parliamentary constituencies